← Back to Skills Marketplace
144
Downloads
0
Stars
0
Active Installs
2
Versions
Install in OpenClaw
/install smartocr
Description
识别车辆证件（行驶证正页/副页）和收据/发票图片，返回结构化 JSON 数据。 支持图片 URL 和本地文件两种方式，需要 API Key。
Usage Guidance
This skill will read your OpenClaw session files to extract images and upload those images (as Base64) to an external API (default: https://smartocr.yunlizhi.cn). Before installing: 1) Confirm you trust the SmartOCR provider and its handling of uploaded images and API keys. 2) Do not use sensitive images (IDs, private documents) until you verify the endpoint and privacy policy. 3) The skill omitted declaring SMARTOCR_API_KEY in its registry metadata—expect to set SMARTOCR_API_KEY in your OpenClaw env; consider using a scoped/test key. 4) If you prefer to avoid external uploads, run your own local SmartOCR instance and set SMARTOCR_API_URL to a localhost address. 5) Because the source/homepage is unknown, prefer running the scripts in a sandbox or reviewing/hosting them yourself rather than granting broad access immediately.
Capability Analysis
Type: OpenClaw Skill
Name: smartocr
Version: 1.0.1
The skill provides OCR functionality for vehicle documents and receipts by sending image data to an external API (smartocr.yunlizhi.cn). It includes a script, `smartocr_from_session.py`, which programmatically accesses OpenClaw's internal session history files in `~/.openclaw/agents/` to extract image data from previous user messages. While these capabilities are aligned with the stated purpose of processing images from the chat, the direct access to session logs and the transmission of potentially sensitive document data to a third-party service are high-risk behaviors.
Capability Assessment
Purpose & Capability
The skill's name/description and the included scripts align with OCR of vehicle documents and receipts. However, registry metadata lists no required environment variables or primary credential while the SKILL.md and both Python scripts require SMARTOCR_API_KEY (and optionally SMARTOCR_API_URL). This mismatch between declared requirements and actual runtime needs is incoherent and should be fixed or explained by the author.
Instruction Scope
The SKILL.md and scripts instruct the agent to read OpenClaw session files (default: ~/.openclaw/agents/{agent}/sessions/*.jsonl) to extract base64 image data and then POST that data to the SmartOCR API. Reading session files is within the stated purpose (to process images sent in chat), but it accesses potentially sensitive conversation history. The skill uploads raw image base64 to an external endpoint (default https://smartocr.yunlizhi.cn), so verify the endpoint and privacy policy before use.
Install Mechanism
There is no external install/download step—this is instruction+script based and only requires python3 and the requests library. No remote archive downloads or unusual install locations are used.
Credentials
At runtime the scripts require SMARTOCR_API_KEY (and optionally SMARTOCR_API_URL and OPENCLAW_HOME). The skill metadata did not declare SMARTOCR_API_KEY as a required env var/primary credential, which is a misleading omission. Requesting an API key to call an external service is proportionate for OCR, but the missing declaration reduces transparency. Also note the API key prefix 'sk-'—verify provider trust and key scope before use.
Persistence & Privilege
The skill does not request always:true and does not attempt to modify other skills or system-wide configs. It reads session files under the user's OpenClaw home for its intended function, which is a significant but expected data access for this feature.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install smartocr - After installation, invoke the skill by name or use
/smartocr - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.1
- Added scripts/smartocr_from_session.py to support recognizing images uploaded in conversations (e.g., Feishu, WebChat, WeCom) by extracting them from session files.
- Updated configuration instructions to align with the new Openclaw CLI and env.vars structure.
- Expanded SKILL.md documentation with usage and options for the new session image recognition script.
- No breaking changes; existing APIs and main script usage remain unchanged.
v1.0.0
Initial release of SmartOCR — vehicle document and receipt/invoice recognition.
- Supports recognition of vehicle license (front/rear) and receipts/invoices from images
- Accepts image URL or local file input, returns structured JSON data
- Requires API Key configuration with option to set custom API URL
- Provides detailed usage instructions, options, and example commands
- Documents response formats, error codes, and operational guidelines
Metadata
Frequently Asked Questions
What is smart ocr?
识别车辆证件（行驶证正页/副页）和收据/发票图片，返回结构化 JSON 数据。 支持图片 URL 和本地文件两种方式，需要 API Key。 It is an AI Agent Skill for Claude Code / OpenClaw, with 144 downloads so far.
How do I install smart ocr?
Run "/install smartocr" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is smart ocr free?
Yes, smart ocr is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does smart ocr support?
smart ocr is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created smart ocr?
It is built and maintained by leeee0923 (@leeee0923); the current version is v1.0.1.
More Skills